Reindexing raises TypeError on Py2.7

[caspervdw]

Code Sample, a copy-pastable example if possible

import pandas as pd
df = pd.DataFrame({'x': [1., 5., 3.]}, index=[0, 2, 4])
df.index.name = 'frame'
df.reset_index('frame')

Problem description

Resetting the index on pandas 0.20.1 and python 2.7 results in the following exception:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "C:\Anaconda3\envs\tpissue\lib\site-packages\pandas\core\frame.py", line 3055, in reset_index
    if level is None or i in level:
TypeError: 'in <string>' requires string as left operand, not int

In the example, it doesn't matter if you use Unicode or non-unicode strings. (using from __future__ import unicode_literals). Ref soft-matter/trackpy#446 @manuescobedo

Expected Output

A reindexed DataFrame.

Output of pd.show_versions()

INSTALLED VERSIONS ------------------ commit: None python: 2.7.13.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 58 Stepping 9, GenuineIntel byteorder: little LC_ALL: None LANG: None LOCALE: None.None

pandas: 0.20.1
pytest: None
pip: 9.0.1
setuptools: 27.2.0
Cython: None
numpy: 1.12.1
scipy: 0.19.1
xarray: None
IPython: None
sphinx: None
patsy: None
dateutil: 2.6.0
pytz: 2017.2
blosc: None
bottleneck: None
tables: 3.4.2
numexpr: 2.6.2
feather: None
matplotlib: 2.0.2
openpyxl: None
xlrd: None
xlwt: None
xlsxwriter: None
lxml: None
bs4: None
html5lib: None
sqlalchemy: None
pymysql: None
psycopg2: None
jinja2: 2.9.5
s3fs: None
pandas_gbq: None
pandas_datareader: None

[gfyoung]

I believe you are incorrectly using the reset_index method. The label you pass in MUST belong to the index of the DataFrame. In the context of your example, that means DataFrame.index.name must be "frame" per the documentation here.

That being said, this error message makes no sense, so we can improve on that so that it won't be as confusing next. Would you like to update / improve it?

[caspervdw]

@gfyoung Thanks for the fast reply. I think I might have simplified the code example too much. Sorry about that. The same exception is there if I execute the following script:

import pandas as pd
df = pd.DataFrame({'x': [1., 5., 3.]}, index=[0, 2, 4])
df.index.name = 'frame'
df.reset_index('frame')

Specifying the name of the only indexer does work on pandas 0.13/py27, 0.16/py27, 0.19/py35, and 0.20.1/py35.

[jreback]

these are as expected. You cannot reset a named index that doesn't exist.

[caspervdw]

@jreback The named index does exist, I made a mistake in my first code example. For clarity, I now edited my first message.

[gfyoung]

@caspervdw : I can't reproduce the error. The code works perfectly fine on Python 2.7 and 3.6 (using 0.20.2 though). Can you double-check this?

[gfyoung]

Also @jreback , I'd like to keep this open, as I would like to see if we can improve the error message

[caspervdw]

Thanks @gfyoung for reopening. I can have a look a the error message for sure, but first I want to see where my issue is coming from.

@manuescobedo brought this up at soft-matter/trackpy#446. We use CI and never encountered this, however with the specific configuration Py27, pandas 0.20.1, I can reproduce. @manuescobedo, could you try to reproduce your error using my minimal example above?

When I have more time I'll try to dive a bit deeper into this. I'll let you know if I find anything.

[gfyoung]

Thanks @jreback for reopening.

Ummm...that was me in fact 😉

Also, try upgrading pandas to 0.20.2, I can't reproduce the issue there. I can however, confirm that that was (we're past that release now) on 0.20.1.

[caspervdw]

I can confirm that the issue disappears on 0.20.2. If I try to reindex using a non-existing index name, I get TypeError: Level name must be same as name '(None)', which is OK I think. So the issue can be closed for my part.